0

I am currently collecting data for a project at University. My aim is essentially to investigate whether actively managed equity-based mutual funds outperform passive mutual funds. I am using the CRSP Survivor-Bias-Free US Mutual Fund Database as the source for my data. However, I am having trouble with sorting the data. I have no experience with MatLab, R or similar coding languages. Therefore, my only solution seems to be sorting the data using the "Conditional statement" feature in wrds and using Excel.

Until now, I have managed to discard all funds that are not equity-based or mixed equity and fixed income. Now I have to exclude all funds that have a front-or rear-load as I only want data on no-load funds. I have tried sorting out load-funds by creating the following conditional statement in my wrds query: "WHERE front_load = 0 AND dollar_amt = -99". This should only include funds with a maximum load of 0, because a missing (-99) "dollar_amt" means that the given front-load is the maximum load for the fund.

The problem I come across is that when comparing the resulting dataset with a dataset that shows all funds and their loads, I notice that some of the funds in my no-load dataset show up with loads in the other dataset. This is because no-load dataset does not take into account that a fund might have been a no-load fund in a given time period, but later has become a fund with load. Therefore, I somehow have to exclude funds that have ever had a fund load.

I am hoping that someone is familiar with the wrds database and the options regarding conditional statements. If not, I understand that it is hard to provide me with a solution.

Thank you.

The database I use: https://wrds-www.wharton.upenn.edu/pages/get-data/center-research-security-prices-crsp/quarterly-update/mutual-funds/

Aleksander
  • 1
  • The WRDS documentation seems to be behind their paywall. You could first get a list of all funds that do have a fund load at any time during your observation period. Then you get the distinct fund ids from that file (use Excel->Data->Remove Duplicates) and put these fund IDs into some exlusion filter on their web based query builder. Is that doable? – Kermittfrog Mar 30 '21 at 14:13
  • @Kermittfrog Thank you very much for the excellent suggestion. Unfortunately, the query builder does not have an exclusion filter. However, I did the following Excel formula which seems to be able to identify all no-load funds: =IF(AND(dollar_amt=-99,front_load=0,crsp_fundno1<>crsp_fundno2),1,""), where crsp_fundno1 is the fund number of the current row and crsp_fundno2 is the fund number of the next row, i.e. two consecutive funds in the dataset. That bit should ensure that I do not include a no-load fund that has changed to a load fund later in the observation period. Once again, thank you – Aleksander Mar 30 '21 at 16:10
  • Hi, yes of course you can identify all in excel and build your exclusion list there. – Kermittfrog Mar 30 '21 at 16:34

0 Answers0